Forwarding on behalf of MSE Prof. Ale Strachan: Thank you, Lisa Stacey Administrative Assistant to the Department Head School of Materials Engineering Neil Armstrong Hall of Engineering o: 765-494-4095 f: 765-494-1204 [3749DD84]<https://urldefense.proofpoint.com/v2/url?u=https-3A__www.purdue.edu_-3Futm-5Fsource-3Dsignature-26utm-5Fmedium-3Demail-26utm-5Fcampaign-3Dpurdue&d=DwMFAg&c=l45AxH-kUV29SRQusp9vYR0n1GycN4_2jInuKy6zbqQ&r=cYZgnbkB39R9SYPCvL_MkXuih1H-IjjKyRBtQtQ6ib8&m=H5ec4nXIY6Mt-FgrmXhcB1ClcZZRWvtSm3BzPca31UE&s=ra1N_tE4A3L7IxFqllrao41XR8k0ynZZYRxh3KMpTwY&e=> ________________________________ Subject: nanoHUB free hands-on workshop on machine learning: Dr. Arun Mannodi-Kanakkithodi, Argonne. Dear Colleagues, nanoHUB is excited to offer another session of our ongoing Hands-on Data Science and Machine Learning Training Series. The series is designed to enhance your skill set with practical tools you can use in your research or teaching. Title: Machine Learning Framework for Impurity Level Prediction in Semiconductors Date/Time: November 11, 2020 / 1 PM – 2 PM EST By: Arun Mannodi-Kanakkithodi, Postdoctoral researcher at Argonne National Laboratory and Incoming Assistant Professor of Materials Engineering at Purdue University Register here (limited seats): https://purdue.webex.com/purdue/onstage/g.php?MTID=e088a6ccfa042d4ac13bdb445... Abstract: In this work, we perform screening of functional atomic impurities in Cd-chalcogenide semiconductors using high-throughput computations and machine learning. High-performance computing resources located at Argonne National Lab (Carbon at CNM and LCRC) and Berkeley Lab (NERSC) were utilized to generate large databases of impurity properties from first principles-based density functional theory (DFT) computations. This dataset was combined with material descriptors—ranging from coordination environments to tabulated elemental properties to cheaper DFT data—to train state-of-the-art regression models. LASSO, random forest, and kernel ridge regression techniques were used and the predictive models were optimized with respect to the type and quantity of training data, optimal hyperparameter sets, and cross-validation errors. The best models thus achieved were deployed to make predictions for the combinatorial chemical space of all possible impurity atoms in Cd-chalogenide compounds, following which screening was performed on the basis of their relative energetics, leading to a shortlist of impurities for every compound which can affect a desirable or disastrous change in the optical and electronic properties of the semiconductor. The data and models developed in this work have major consequences for semiconductor applications ranging from solar cells to infrared sensors to quantum information sciences. Publication: https://www.nature.com/articles/s41524-020-0296-7. Bio: Arun Mannodi-Kanakkithodi is a computational materials scientist working as a postdoctoral researcher at the Center for Nanoscale Materials at Argonne. His research interests include AI-driven design of novel materials for energy applications, optimizing opto-electronic properties of novel semiconductor compositions, generating and mining computational materials databases, materials informatics, and data-driven materials discovery. Arun will begin a new position as assistant professor in Materials Engineering at Purdue University in December 2020. Additional information. Our series is aimed at active researchers and educators and little prior coding experience is required. In addition, all exercises will use nanoHUB cloud computing resources, no need to download or install any software. All you need is an internet connection and a browser. After the training sessions, you will be able to continue using nanoHUB for your research or class. You can find additional information and sign up links at https://nanohub.org/groups/ml/handsontraining. Here you can also find recorded material of our previous workshops and refresh your knowledge of Jupyter notebooks, querying databases, and training basic machine learning models such as neural networks for classification and regression. We hope that you can attend this workshop and walk away with enough information and practical skills to kickstart your foray into deep learning methods for your research or classroom use. Best regards, Prof. Ale Strachan, Materials Engineering, Purdue University Deputy Director, nanoHUB. http://nanohub.org/ -- Alejandro Strachan Professor of Materials Engineering, Purdue University Network for Computational Nanotechnology - nanohub.org<http://nanohub.org> Center for Predictive Materials and Devices (c-PRIMED)
